Microfluidic Devices for Blood Fractionation
The MIT Faculty has made this article openly available. Please share
how this access benefits you. Your story matters.
Citation
Hou, Han Wei, Ali Asgar S. Bhagat, Wong Cheng Lee, Sha
Huang, Jongyoon Han, and Chwee Teck Lim. “Microfluidic
Devices for Blood Fractionation.” Micromachines 2, no. 4 (July
20, 2011): 319–343.
As Published
http://dx.doi.org/10.3390/mi2030319
Publisher
MDPI AG
Version
Final published version
Accessed
Thu May 26 22:52:26 EDT 2016
Citable Link
http://hdl.handle.net/1721.1/88055
Terms of Use
Creative Commons Attribution
Detailed Terms
http://creativecommons.org/licenses/by/3.0/
Micromachines 2011, 2, 319-343; doi:10.3390/mi2030319
OPEN ACCESS
micromachines
ISSN 2072-666X
www.mdpi.com/journal/micromachines
Article
Microfluidic Devices for Blood Fractionation
Han Wei Hou 1,2, Ali Asgar S. Bhagat 1, Wong Cheng Lee 1,3, Sha Huang 4, Jongyoon Han 1,4,5
and Chwee Teck Lim 1,2,3,6,7,*
1
2
3
4
5
6
7
BioSystems and Micromechanics (BioSyM) IRG, Singapore-MIT Alliance for Research and
Technology (SMART) Centre, Singapore 117543; E-Mails: hwhou@nus.edu.sg (H.W.H.);
ali.bhagat@gmail.com (A.A.S.B.); wclee@nus.edu.sg (W.C.L.)
Division of Bioengineering, National University of Singapore, Singapore 117576
NUS Graduate School for Integrative Sciences and Engineering, National University of Singapore,
Singapore 117456
Department of Electrical Engineering & Computer Science, Massachusetts Institute of Technology,
Cambridge, MA 02139, USA; E-Mails: huangsha@mit.edu (S.H.); jyhan@mit.edu (J.H.)
Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge,
MA 02139, USA
Department of Mechanical Engineering, National University of Singapore, Singapore 117576
Mechanobiology Institute, Singapore 117411
* Author to whom correspondence should be addressed; E-Mail: ctlim@nus.edu.sg;
Tel.: +65-6516-6564; Fax: +65-6772-2205.
Received: 11 May 2011; in revised form: 27 June 2011 / Accepted: 6 July 2011 /
Published: 20 July 2011
Abstract: Blood, a complex biological fluid, comprises 45% cellular components
suspended in protein rich plasma. These different hematologic components perform
distinct functions in vivo and thus the ability to efficiently fractionate blood into its
individual components has innumerable applications in both clinical diagnosis and
biological research. Yet, processing blood is not trivial. In the past decade, a flurry of
new microfluidic based technologies has emerged to address this compelling problem.
Microfluidics is an attractive solution for this application leveraging its numerous
advantages to process clinical blood samples. This paper reviews the various microfluidic
approaches realized to successfully fractionate one or more blood components.
Techniques to separate plasma from hematologic cellular components as well as isolating
blood cells of interest including certain rare cells are discussed. Comparisons based on
common separation metrics including efficiency (sensitivity), purity (selectivity), and
Micromachines 2011, 2
320
throughput will be presented. Finally, we will provide insights into the challenges associated
with blood-based separation systems towards realizing true point-of-care (POC) devices and
provide future perspectives.
Keywords: blood separation; microfluidics; cell separation; enrichment; disease detection
and diagnosis
1. Introduction
Blood, perhaps the most important biological fluid, performs many fundamental functions to
maintain homeostasis; from transporting nutrients and oxygen to tissues and organs to regulating pH
and temperature. It also provides an efficient transit system through the vascular network for transport
of immune cells for defense against foreign microbes and wound healing. As blood contains a myriad
of information about the functioning of the human body, complete blood analysis has been a primary
diagnostic test in our healthcare system. Broadly, plasma and cells constitute the two main blood
components each with ~60% and 40% volume fractions respectively. The blood cellular components
mainly consist of red blood cells (RBCs), leukocytes and platelets. Other rare-cells including fetal cells
during pregnancy and epithelial tumor cells in cancer patients may also be found circulating in blood.
Of the ~5 billion cells per milliliter of blood, red blood cells (RBCs) make up >99% of all cellular
components. Due to the vast majority of RBCs, fractionating the various target components from blood
has been a challenging problem both from a medical and engineering perspective.
Conventional plasma and leukocyte separation methods have relied on membrane-based filtration.
However, the high cellular fractions lead to membrane clogging and compromise separation efficiency.
To address this, enormous progress has been made in the past few years in the area of microfluidics
based blood component separations. Microfluidics is the science of studying fluid flow behavior at the
microscale and the development of miniaturized analysis systems that take advantage of the unique
physics emergent at these small scales [1,2]. Microfluidics leverages its many distinct advantages such
as low sample volume for processing blood, as these samples are precious and rare especially in the
case of neonatal care. A comprehensive review of processing blood on a microchip discussing the
various sample preparation challenges, cell separation methods and integrated blood analysis modules
was presented by Toner and Irimia [3]. However, the field of microfluidics based blood fractionation
has significantly advanced since then and a review of the various microfluidic approaches developed in
the past few years is warranted.
In this paper we review recent microfluidic methods realized to successfully fractionate one or more
blood components. Although microfluidic based cell separation techniques have been extensively
reviewed in the past, to our knowledge this is the first attempt to summarize microfluidics based
techniques to separate the individual blood components [4-9]. Detailed description of the various
separation principles for each component will be discussed and comparisons based on separation
metrics including efficiency (sensitivity), purity (selectivity), and throughput will be drawn. The
readers are encouraged to read the original papers for additional details. Analysis of these individual
components for clinical diagnosis or fundamental studies is beyond the scope of this review.
Micromachines 2011, 2
321
2. Plasma Separation
Many clinical tests and research assays frequently require the removal of blood cells from plasma
for downstream analysis and diagnosis. A simple diagnostic test of blood plasma for biological
substances such as cholesterol, glucose and lactate can be used to evaluate the general state of health of
various internal organs. The quantification of individual plasma proteins is also routine and important
for diagnosis and monitoring of many diseases including paraproteinaemias, hemoglobinopathies, and
genetic abnormalities [10,11]. However, it is critical to remove the large volume of interfering blood
cells from whole blood so that the purified plasma can be assessed and diagnosed accurately.
Traditionally, plasma is extracted from blood in laboratories and clinics by centrifugation with
conventional bench-top centrifuges which are known to be expensive, time consuming and labor
intensive. The practice of centrifugation has also been realized on the microscale using compact
disk-like platforms with manifolds and a spinning motor plate to achieve centrifugal pumping, fostered
by the need to develop an integrated total analysis system [12,13]. Haeberle and coworkers designed a
decanting structure on a rotating disk for the extraction of plasma [12]. The technique efficiently extracts
2 µL plasma from a 5 µL blood sample rotated at 40 Hz frequency within 20 s. The extracted plasma
contained <0.11% blood cell concentration independent of rotation frequency. Recently, an innovative
method to isolate plasma for resource limited settings has been demonstrated by Wong et al. using an
egg beater as a hand-powered centrifuge [Figure 1(A)] [14]. By spinning the paddle, human whole blood
loaded into polyethylene (PE) tubing attached to the end of the paddle of the egg beater is centrifuged.
The blood cells settle down at the distal end of the PE tubing while plasma extracted from the
supernatant can be used to perform diagnostic assays. The technique extracts ~58 µL plasma from
100 μL whole blood after 8 min of centrifugation with very high purity (cell concentration
~5 × 104 mL−1). However, during centrifugation the sedimented blood cells can easily lyse (RBCs
hemolysis), thereby releasing intracellular components which contaminate the plasma sample [14]. To
overcome this limitation, several microfluidic techniques for plasma isolation from blood have been
recently reported.
Perhaps the most common way of separating plasma from blood cells using microfluidic devices is
based on the principle of particle retention. These devices frequently involve the use of micromachined
filters or meshes to retard or block blood cell movement, thereby allowing the collection of the plasma
portion of blood. In particular, comb and weir-type filter structures with well defined geometries have
been designed and demonstrated [15-17]. Designing weirs with a gap size of 0.5 µm, VanDelinder and
Groisman reported a cross-flow based microfluidic device for continuous plasma separation [16]. The
0.5 µm gap permits the flow of plasma while impeding the RBCs, leukocytes and platelets. The device
is capable of operating continuously for an hour extracting ~8% plasma at 0.65 µL/min throughput and
<0.1% hemolysis. Similar design employing 0.5 µm gap transverse-flow microfilters was also reported
by Crowley and Pizziconi [17]. Size-exclusion based techniques relying on integrated microporous
fleece or membranes have also been developed [18,19]. Moorthy and Beebe fabricated porous filters
inside a microchannel using emulsion photo-polymerization with the porosity of the filter being a
function of the pre-polymer mixture constituting 2-hydroxyethyl methacrylate (HEMA, monomer),
ethylene glycol dimethacrylate (EGDMA, cross-linker), and 2,2-dimethoxy-2-phenyl-acetophenone
(DMPA, photo-initiator) in water. The device was tested with diluted blood (1:20) to demonstrate
Micromachines 2011, 2
322
separation. However, both the weir-based and membrane-based devices frequently suffer from RBCs
contamination due to their extremely high deformability. This also poses enormous challenge to the
fabrication as very small pore size or filters on the order of 1 μm or less are required to retain the
RBCs effectively. Moreover, as the geometrical features of these devices become smaller, clogging
becomes a major concern and blood has to be diluted before separation thereby compromising
throughput and operation time. Another novel method to separate plasma from whole blood using
bead-packed microchannels acting as filters has been demonstrated recently [Figure 1(B)] [20]. Blood
introduced into the channel is driven by the capillary force induced by the bead column. As the RBCs
movement is impeded by the packed beads, cell-free plasma is driven through the bead column.
Without the use of microfilters, this technique minimizes RBCs clogging, but is unsuitable for rapid
clinical diagnostics due to long separation time as a result of slow capillary action.
In order to circumvent the aforementioned problems, several groups have investigated RBC
depletion using flow through devices which function solely on inherent hydrodynamic separation
mechanisms. Separating blood plasma based on the inherent hydrodynamic separation mechanisms
frequently rely on the Zweifach-Fung effect, also known as bifurcation law [21-23]. The technique
exploits the tendency of cells at branching points of microchannels to flow into the wider channel
having a higher flow rate. By varying the channel widths of the daughter channels, Yang et al. reported
a minimum flow rate ratio of 6:1 required between the main and smaller branching channels to obtain
successful blood plasma separation [Figure 1(C)] [21]. The plasma selectivity obtained was close to
100% and was independent of the inlet hematocrit levels while the collected plasma volume varied
from 15 to 25% depending on the inlet hematocrit levels. Plasma separation from blood cells has also
been achieved using “plasma skimming” or the Fahraeus effect. In microchannels flow under
Poiseuille conditions, the highly deformable RBCs migrate towards the axial centre of the
microchannels parallel to flow. This creates a lower cell concentration or “cell free” region near the
channel walls where plasma can be extracted easily. Jaggi et al. demonstrated the use of a high-aspect
ratio microchannel device with many vertical channel bifurcations arranged in parallel to extract
plasma from blood [24]. The device permits very high throughput operating at relatively high flow
rates of between 2 to 5 mL/min with high separation (cell depletion) efficiency (~90%) at 4.5%
hematocrit. However, the cell depletion efficiency drops significantly to ~30% when tested with
physiological hematocrit (45%). Blattert et al. combined centrifugal forces with plasma skimming to
demonstrate plasma extraction using a simple microchannel bend [25,26]. The centrifugal forces
results in differing settling velocities between the cells and plasma resulting in the collection of cell
free plasma near the inner wall of the microchannel following the bend region. As this technique
also relies on plasma skimming, blood dilution is necessary as a ~90% cell depletion efficiency at
~5% hematocrit reduces to ~40% at 45% hematocrit.
Micromachines 2011, 2
Figure 1. Plasma separation from blood. (A) left: Schematic diagram shows the
centrifugation based separation principle using an egg beater with the blood-filled PE
tubing attached to the paddle (inset illustrates the photograph of a manual hand-powered
egg beater). centre: Phase contrast images showing RBCs concentrations in the plasma
layer and cell pellet region of the PE tubing after 10 min spinning. right: Graph showing the
plasma volume generated against centrifugation time from a 100 µL blood sample. Reprinted
with permission from [14] copyright 2008, The Royal Society of Chemistry; (B) left:
Graphical illustration of plasma separation from whole blood using icrobeads-packed
microchannel. As blood flows through the hydrophilic channel by capillary forces, RBCs
movement is impeded by the packed beads while the plasma (yellow) continues to move
forward. right: Photograph showing the formation of four separated regions as blood flows
through the bead-packed channel corresponding to different stages of plasma separation.
Figure reproduced from [20], copyright 2010, with kind permission from Springer
Science + Business Media, LLC; (C) Plasma separation using hydrodynamics effects. left:
Schematic representation of the Zweifach–Fung effect in which RBCs travel into the daughter
channel with higher flow rate when the flow rate ratio is more than 2.5 times compared to the
smaller daughter channel. The image at the bottom shows the schematic of the actual plasma
separation device with the plasma separation region consisting of 5 bifurcated plasma
skimming channels. centre: Photograph of the blood plasma separation region at a flow rate
of 10 µL/h. right: Comparison between experimental and computational fluid dynamics
(CFD) simulation results of the total plasma separation volume percent with respect to the
upstream microchannel hematocrit level. The numbers indicate at each data point represent
the inlet hematocrit levels. Redrawn with permission from [21], copyright 2006, The Royal
Society of Chemistry.
323
Micromachines 2011, 2
324
Figure 2. (A) Plasma separation using acoustophoresis. Cross-sectional schematics of a
microchannel excited by a λ/2 wavelength ultrasonic standing wave indicating the focusing
of the RBCs at the channel center by the primary acoustic radiation force. By designing a
multiple outlet system enriched blood cell fractions are removed from the center of the
channel while clean plasma fraction can be withdrawn from the side. A transducer placed
underneath the chip generates an ultrasonic standing wave between the channel walls
perpendicular to the flow. Reprinted with permission from [27], copyright 2009, American
Chemical Society; (B) Plasma separation using dielectrophoresis (DEP). The image shows
the top and cross sectional views of plasma extraction principle from diluted blood (1:9) in
the presence of electric field gradient. The strong DEP forces traps ~97% of RBCs at the
electrodes extracting ~300 nL of plasma from 5 µL of blood. Figure reprinted from [33],
copyright 2010, with permission from Elsevier; (C) left Schematic illustrating the principle
of inducing swirling flows in microchambers by bulk electrohydrodynamics thrust or ionic
winds generated by sharp corona electrode tip. right Successive time-lapse images
indicating the concentration of RBCs near the bottom of the microfluidic chamber toward a
stagnation point in a 0.4% hematocrit sample at 286 kV/m and 60 kHz field. Figure
reproduced with permission from [31], copyright 2007, American Institute of Physics.
Plasma separation techniques from blood using active energy sources such as acoustophoresis,
electrohydrodynamics and dielectrophoresis have also been explored recently [27-33]. Acoustic forces
generated by ultrasonic standing waves can be used for plasma separation in microchannels. The
radiation force generated on the cells by these acoustic waves moves them into pressure nodes of the
standing wave field. By exciting the microfluidic channel with half wavelength width ultrasonic wave,
the cells/particles can be focused at the pressure node located at the center of the channel. Thus,
pumping blood through the channels, the erythrocytes can be focused at the channel center and
removed from the central outlet while the purified plasma can be continuously collected at the side outlets
[Figure 2(A)] [27-29]. The device is capable of generating pure fractions of plasma (<6 × 106 cells/mL)
Micromachines 2011, 2
325
at ~80 μL/min flow rate. Dielectrophoresis (DEP) offers another effective method of separating blood
plasma relying on an inhomogeneous electric field between two electrodes to induce electrostatic
forces [Figure 2(B)] to trap cells. When polarizable cells (RBCs) are placed in the non-uniform electric
field, a net electric force is imparted on the cells due to an induced or permanent dipole. Separation of
blood plasma from diluted blood (1:9) using a DEP microfluidic device was demonstrated by
Nakashima et al. [33]. Blood is injected into the main channel by capillary force and in the presence of
an electric field the RBCs are trapped at the electrodes while cell free plasma is collected downstream.
Applying an AC signal of 10 V and 10 MHz, the device is capable of removing about 97% of blood
cells to generate 300 nL plasma from 5 μL of blood [33]. Recently, an interesting plasma separation
technique based on creating swirling flows in a microfluidic chamber induced by bulk
electrohydrodynamics thrust or ionic winds has been reported [Figure 2(C)] [30,31]. Applying a
potential that exceeds the threshold ionization voltage for air across a mounted sharp corona electrode
tip creates a nearly singular electric field giving rise to a corona discharge. The bulk
electrohydrodynamic air thrust generated by the transfer of momentum from collision between the
counter-ions with air molecules results in interfacial shear leading to recirculating flows [Figure 2(C)].
The cells suspended in the liquid are thus concentrated towards the bottom of the chamber by these
inward radial forces. Using this technique, Yeo and co-workers successfully isolated plasma from
RBCs within 200 s applying 0.94 kV voltage and 90 kHz frequency. The separation is efficient with
the separated plasma consisting of <0.003% hematocrit cell contamination.
3. Leukocyte Separation
Leukocytes (or white blood cells (WBCs)) are the nucleated fraction of blood cells, mainly
consisting of neutrophils, basophils, eosinophils, monocytes and lymphocytes, and are primarily
responsible for all immunological functions. Compared to the vast abundance of erythrocytes
(~5 million cells/μL), leukocytes are present at a much lower concentration, (typically between 4,000
to 11,000 cells/μL) constituting ~0.1% of all blood cells. Isolation of leukocytes is essential for many
hematological analysis and clinical diagnostic tests for monitoring disease progression [34]. Removal of
leukocytes from blood, leukapheresis, is essential to reduce the white blood cell count in treating leukemic
patients as well as to retrieve blood stem cells in patient with hematologic malignancies [35,36]. It is also
essential to remove all leukocytes before performing a blood transfusion to minimize transmission of
cell-associated infectious agents (cytomegalovirus, herpes virus, human T-cell lymphoma virus), graft
rejection rates and prevent febrile transfusion reactions [37].
Traditionally, these cells are separated from blood by centrifugation exploiting the differences in the
buoyant gravity by using a density gradient solution (Ficoll-Paque™). Besides density, leukocytes and
RBCs are also structurally different both in terms of shape and size. Leukocytes are spherical
nucleated cells with typical diameters ranging from 6 µm to 10 μm [38-40]. Red blood cells (RBCs) on
the other hand have a biconcave disc-like structure with an average diameter of 7–8 µm and 2–3 µm
thickness. Due to the high deformability of RBCs aided by the absence of nucleus, membrane-based
filters are also commonly employed to filter leukocytes from blood [41,42]. The inherent differences in
size and deformability between the RBCs and leukocytes have led to the development of several
microfluidics techniques based on the principles of cell retention and retarded cell movement [40,43-45].
Micromachines 2011, 2
326
The use of pillars in particular is an attractive technique for size and deformability based cell sorting in
microfluidic devices. For example, Panaro et al. designed a silicon microchip with rectangular pillars
spaced 3.5 µm to separate leukocytes from whole blood for subsequent PCR reactions [43]. The
“deterministic lateral displacement” (DLD) or the “bump array” microdevice developed by Huang et al.
has also been applied for blood fractionation [Figure 3(A)] [46,47]. The microfluidic device design
consists of arrays of micro-pillar structures placed within the main flow channel leading to the
formation of multiple cell streams based on size. Cells larger than the critical diameter dc = 20% × 2w
(w is the separation gap between two adjacent pillars) follow a deterministic path while smaller cells
remain unperturbed moving in an average downward flow direction. The bump array was successfully
applied for fractionating blood components by tailoring the design to match the sizes of the blood cells.
Apart from achieving 110-fold leukocyte enrichment, the device was capable of fractionating the
individual WBC components. However, a processing flow rate of 1 µL/h results in an extremely low
device throughput.
A comprehensive study conducted by Ji and coworkers investigated the performance of four main
types of silicon microfilter designs, namely, weir-based, membrane, pillars and cross-flow for efficient
leukocyte trapping [44]. The critical dimension in all the four microfilter designs was fixed to 3.5 µm.
Their results confirm the superiority of the pillars and cross-flow devices for leukocyte trapping with
>60% of leukocytes trapped in the device. However, the cross-flow based device outperforms other filter
designs in its blood handling capacity (>300 µL) and allowing >50% of RBCs to pass through. Similar
cross-flow based leukocyte separation device have also been reported by other researchers. [40,45].
Sethu et al. developed a diffusive filter design consisting of a three parallel channels for leukocyte
removal [Figure 3(B)] [40]. The two side channels are connected to the main channel through 2.5 µm
deep microfabricated sieves. The sieves allow the passage of >50% deformable RBCs to the side
channels while retaining >97% of the larger leukocytes in the center main channel at 5 µL/min
throughput. Recently, VanDelinder and Groisman described a technique to separate cells by size and
exchange carrier medium by perfusion in a continuous cross-flow based microfluidic device [45]. By
designing 3 µm shallow connecting channels to impede leukocyte migration, the device retains >98%
leukocytes in the center channel achieving a remarkably high enrichment factor of 4,000 fold by
effectively depleting all RBCs.
While all the aforementioned techniques effectively isolate leukocytes from blood, these techniques
commonly suffer from clogging issues leading to shorter operational times and lower throughput. To
address these limitations, novel separation techniques taking advantage of the emergent hydrodynamic
behavior at the microscale have been developed. Hydrodynamic filtration methods relying on the
relative position of the center of the cells in microchannels have been recently demonstrated for
separating leukocytes from RBCs [Figure 4(A)] [48,49]. By designing multiple side branching
channels, particles/cells of varying diameter can be aligned along the channel sidewalls and collected
at different outlets downstream [48]. The device reports a ~29 fold leukocyte enrichment at a relatively
high flow rate of 20 µL/min with no clogging. Similarly, the device developed by Zheng et al. takes
advantage of the position of the leukocytes and RBCs along the flow streamlines to separate them in
different outlets [49]. As the center of the leukocytes is further away from the channel sidewalls than
the smaller RBCs, this allows the leukocytes to be extracted at outlets placed further downstream with
a high separation efficiency of >97%. Another novel separation technique based on the principle of
Micromachines 2011, 2
327
hydrophoresis has been reported for leukocyte isolation in microchannels patterned with slanted
grooves on the top and bottom channel walls [50]. The hydrodynamic interaction of cells with a fluid
flow induced by slant obstacles forces them to flow in different trajectories depending on the cell
diameter. Using diluted blood (1:20), the device demonstrates a 210 fold leukocyte enrichment with
85% recovery at a processing throughput of 4,000 cells/s.
Figure 3. Leukocytes separation from blood. (A) left: Schematic of the fractionating
device used for separating blood components based on deterministic lateral displacement
(DLD) technique. centre Graphic illustrating the DLD separation principle together with
histogram showing concentrations of different blood leukocyte components collected at the
three outlets after separation. right: Experimental images showing the displacement of the
bigger leukocytes (labeled with Hoechst 33342 dye) away from the inlet sample stream
towards the right side of the channel by the bump array. Reprinted with permission
from [47], copyright 2006 National Academy of Sciences, USA; (B) left: Schematic of the
diffusive filter based microdevice for continuous flow fractionation of leukocytes from
whole blood based on size and shape differences. The microchannel design consists of
three parallel channels connected by 2.5 µm deep micro sieves. Blood pumped through the
main channel center is depleted of RBCs at the side channels by cross-flow filtration while
the bigger leukocytes are retained at the centre channel. right: Histogram showing the
experimental fractionation results as a function of flow rate. At an optimal flow rate of
5 µL/min >97% of leukocytes are collected from the center outlet with 50% RBC depletion.
Redrawn with permission from [40], copyright 2006, The Royal Society of Chemistry.
Micromachines 2011, 2
Figure 4. (A) top: Leukocyte separation using hydrodynamic filtration. Schematic
diagrams showing the particle position and trajectory at a branch point for low, medium,
and high relative flow rates between the main and side channels. As liquid is continuously
withdrawn from the side channels, all particles concentrate along the channel sidewalls and
are separated into multiple side channels based on size. bottom: Microscopic images
showing enrichment of leukocytes (indicated by arrows) before filtration (a) and at outlet 1
and 3 after second filtration (b). Figure reproduced with permission from [48], copyright
2005, The Royal Society of Chemistry; (B) Biomimetic leukocyte separation technique
based on in vivo leukocyte margination phenomenon. top: The larger WBCs (red arrows)
are displaced to the channel sidewalls due to axial migration of RBCs and extracted using
an asymmetrically bifurcated outlet as shown. bottom: Histogram showing the leukocyte
distribution at the end of a 5.5 mm long rectangular microchannel indicating the successful
margination of all the leukocytes towards the channel side accomplishing a 34-fold
enrichment. Figure redrawn with permission from [51], copyright 2005, American
Chemical Society; (C) Affinity capture of leukocytes using patterned microstructures
based on leukocyte recruitment and rolling. left: Illustration of different stages of cell
adhesion in the square and offset micropillar array. As cells adhere on the lateral surface of
the E selectin coated pillars (R), the fluid shear in the tangential direction make the cells
roll along the pillars. Reaching the end of the pillars, these cells tend to tether and detach
from the surface before being captured again by an adjacent pillar. Insets (red box) shows
the microchannel image with square and offset arrays. centre: Images showing the
accumulation of HL-60 cells captured along the square array at the start (top) and 2 min
later (bottom). right: Graph showing the enrichment of HL-60 cells with time. HL-60 cells
were enriched ~400 times in the square array and ~240 times in the offset array. Figure
reprinted with kind permission from [53], copyright 2005, The Royal Society of Chemistry.
328
Micromachines 2011, 2
329
In vivo physiological phenomenon such as leukocyte margination and recruitment in the
microcirculation has also been mimicked in microfluidic devices for leukocyte separation [51-53]. In
blood vessels with luminal diameter less than 300 μm, RBCs due to their smaller size and higher
deformability than the leukocytes, tend to migrate to the axial centre of the vessel under Poiseuille
flow conditions. As these RBCs migrate towards the axial centre, mechanical collisions between the
leukocytes and the migrating RBCs results in the displacement of the larger leukocytes towards the
vessel wall; a phenomenon termed as margination [52,54]. This bio-mimetic phenomenon can be
exploited to achieve ~34 fold leukocyte enrichment by designing appropriately branched daughter
channels to continuously filter all the leukocytes present along the microchannel sidewalls from
blood [Figure 4(B)] [51]. Recently, separation of leukemic HL-60 and U-937 cells based on the
interaction between the cells and protein coated microchannel surfaces has also been demonstrated
[Figure 4(C)] [53]. By coating pillar surfaces with E-selectin adhesion proteins to mimic in vivo
phenomenon of leukocyte recruitment, the leukemic cells are immuno-captured and arrested from the
passing flow. Due to the high selectivity, this technique permits the capture of >95% cells with
150-fold average enrichment at a throughput of 15,000 cells/min.
Most of the above mentioned techniques separate leukocytes based on their mechanical properties
such as size and deformability. Besides taking advantage of the differences in these distinct mechanical
properties to sort leukocytes from RBCs, the varying surface properties and net charges of these cells
can also be used for efficient sorting as well. For example, SooHoo and Walker developed a novel
method to separate leukocytes from whole blood using microfluidic aqueous two-phase system
(ATPS) [Figure 5(A)] [55]. Using a three inlet system, polyethylene glycol (PEG) and dextran (Dex)
solution are used to create one or two immiscible interfaces inside the microchannel. Blood diluted in
PEG (1:50) is pumped through the center inlet, with dextran solutions pumped through the two side
inlets. The PEG-Dex interface exhibits a force on the cells due to the ATPS interface surface energy.
The magnitude of this force is proportional to the surface area and thus the leukocytes having a larger
surface area experience higher force and prefer being at the interface while the RBCs migrate into the
dextran solution. This simple separation method can process diluted blood at 1 µL/min flowrate with a
~10 fold leukocyte enrichment with a two interface design [55]. Another interesting leukocyte
recovery method taking advantage of the difference in lysing time between the leukocytes and RBCs
when mixed in a hypotonic solution (de-ionized water) has been explored. When mixed with de-ionized
water, RBC lysis completes within 8 s while leukocyte lysis begins after 3 s [56]. By designing a
3-inlet microfluidic cassette (side inlets pump de-ionized water) integrated with a micromixer for
efficient mixing and controlling the exposure time to within 10 s, all leukocytes are recovered along
with the lysed RBC debris at the outlet [56].
Micromachines 2011, 2
Figure 5. (A) left: Schematic illustration of leukocyte separation based on microfluidic
aqueous two-phase system (ATPS). Whole blood diluted in PEG is exposed to either a one
or two PEG-Dex interfaces, represented by the dashed line. Due to differences in surface
energies, the leukocytes prefer the interface while RBCs migrate into the Dex region.
right: Experimental images showing leukocyte separation from blood in one interface
PEG-PEG-Dex system (a-b) and two interfaces Dex-PEG-Dex system (c-d). Figure
reprinted from [55], copyright 2009, with permission from Springer Science + Business
Media, LLC; (B) left: Schematic representation of lateral driven continuous dielectrophoretic
separators using an (a) divergent and (b) convergent type interdigitated electrode pattern.
right: Histograms showing the relative separation percentage of RBCs and WBCs at each
outlet using the divergent (top) and convergent (bottom) type DEP microseparators. Figure
reproduced with permission from [57], copyright 2008, The Royal Society of Chemistry;
(C) top: Diagram showing various forces acting on a magnetically labeled cell passing over
a magnetic stripe. By incubating blood with super paramagnetic CD45 microbeads, the
flow trajectory of leukocytes can be altered by the magnetic stripes. bottom: Time lapse
image showing a single tagged fluorescent leukocyte tracking along a magnetic stripe
inclined at an angle of 9.6° to the fluid flow, indicated by white arrows. The RBCs in blood
continue to remain in the sample stream and are unaffected by the magnetic force. Figure
reproduced with permission from [58], copyright 2004, American Institute of Physics.
330
Micromachines 2011, 2
331
Leukocyte separation using active sorting techniques including dielectrophoresis and magnetic
separation have been reported recently on the microscale. Han et al. developed lateral-driven
dielectrophoretic (DEP) microseparators to isolate leukocytes from blood suspended in a highly
conductive medium [Figure 5(B)] [57]. By subjecting the blood cells to a constant dielectrophoretic
forces generated using patterned convergent and divergent type interdigitated microelectrodes, the
leukocytes were separated from RBCs. As the RBCs have a higher density compared to leukocytes
(RBCs ~ 1,130 kg.m−3, WBCs ~ 1,050–1,080 kg.m−3), the DEP force acting on them is higher
resulting in a larger deflection at the electrodes. The authors report a 92% leukocyte recovery with ~7
fold enrichment at a continuous flow rate of 50 µL/h. Finally, magnetic based leukocyte separation
methods exploiting either inherent differences in the cells magnetic susceptibility or selectively
immuno-labeling magnetic beads have been demonstrated [58-61]. For example, whole blood
incubated with CD45 labeled super paramagnetic microbeads pumped through a microchannel with
patterned magnetic stripes has been explored for leukocyte isolation [Figure 5(C)] [58]. As all
leukocytes express CD45 surface markers to some degree, the magnetic stripes (magnetized by an
externally applied field of 0.08T) alter the flow direction of the CD45 tagged leukocytes from
unlabeled cells, thus achieving separation. Although, the selectivity of this technique is very high due
to immuno-labeling, a significant portion (~50%) of leukocytes were lost either by being permanently
stuck to the magnetic strips or were not sufficiently magnetized to alter their direction [59].
Additionally, magnetic bead sorting often requires pre-treatment of cells, which may damage cells
inadvertently. To overcome this problem, separation of RBCs from leukocytes based on the inherent
magnetic susceptibility of these cells has also been reported [60,61]. By patterning a ferromagnetic
wire within a microchannel the wire concentrates and attracts the paramagnetic RBCs along its length
while the diamagnetic leukocytes are separated in a continuous and efficient manner [61]. Using a
three-stage cascaded system, the authors report a 97% leukocyte recovery with 15× enrichment at
moderate flow rates of 2.5–20 µL/h [61].
4. Platelet Separation
Platelets are the smallest type of cells found in blood (~2–3 µm diameter), and they are involved in
many fundamental physiological processes such as hemostasis and thrombosis [62]. When bleeding
from a wound occurs, circulating platelets adhere and aggregate at the injury site to initiate thrombus
formation until the bleeding is stopped. Patients undergoing general surgery, organ transplants and
trauma treatments frequently require platelet transfusion [63]. Platelets are usually extracted from
blood via centrifugation and by other methods which include platelet-rich plasma preparation (PRP),
the buffy-coat preparation (BC) and apheresis [64]. However, these techniques are inefficient because
a significant portion of inactivated platelets are often lost either through high mechanical shear
stress [65] or the purification process directly [64]. Thus new efforts in developing efficient platelets
extraction are leveraging on existing low stress microfluidic technologies.
Platelet separation in microfluidic systems has mainly been explored with active separation
techniques using acoustophoresis [66] and dielectrophoresis [67]. Platelets can be separated
into multiple fractions in a continuous mode via free flow acoustophoresis (FFA) [Figure 6(A)] [66].
A half-wavelength acoustic standing wave generated between the microchannel side walls applies an
Micromachines 2011, 2
332
acoustic radiation force perpendicular to the direction of the flow. By using a combination of FFA and
medium density manipulation, high separation efficiency can be achieved with 92% of the erythrocytes
ending in the first fraction (center of the channel) while 99% of the platelets collected in the second.
Similarly, by exploiting the fact that platelets are the smallest cells in blood and therefore experiences less
dielectrophoretic force, dielectrophoresis can be used to extract high purity platelets (~95%) from diluted
whole blood at a throughput of ~2.2 × 104 cells/s with minimal platelet activation [Figure 6(B)] [67].
Recently, single-step separation of individual platelets from whole blood by micropatterning of
platelet-specific protein surfaces on solid substrates has been reported using an interfacial platelet
cytometry (iPC) chip [62]. Platelets were separated from whole blood using iPC coated with
fibrinogen, von Willebrand factor (VWF), and anti-CD42b antibody printed “spots” ranging from 2–24 μm
diameter. The paper reports on tests conducted with blood samples from healthy donors and a
Glanzmann’s Thrombasthenia patient. As the platelets are allowed to adhere to the surface, the iPC
chip enables physiologically relevant platelet bioassays based on platelet/protein-matrix interactions.
Figure 6. Platelets separation from blood. (A) top: Illustration of free flow acoustophoresis
(FFA) showing size-based separation of particles or cells into multiple fractions in a
continuous flow manner. The varying acoustic forces equilibrate the larger cells closer to
the channel center and smaller cells farther from the center. bottom: The fractionated
particles can be collected separately by designing five consecutive outlets/fractions. The
histogram shows the separation of platelets, RBCs and leukocytes at the various outlets
indicating majority of platelets collected at outlet 4. Reprinted with kind permission from [66],
copyright 2007, American Chemical Society; (B) top: Device architecture of the two-stage
DEP-activated cell sorter (DACS) chip. The integration of two tandem purification stages
increases the purity of the platelet population at the collection outlet. bottom: Schematic
illustration of the DEP separation principle indicating the forces acting on the blood cells.
The larger blood cells experience a higher DEP force from the electrodes and are deflected
from the side to the central buffer stream while the smaller platelets are a retained in the
side buffer stream. Figure adapted with permission from [67]. copyright 2008 Wiley-VCH
Verlag GmbH and Co. KGaA.
Micromachines 2011, 2
333
5. Rare-Cell Separation
Besides blood constituents, other cells of interest such as circulating tumor cells (CTCs), nucleated
RBCs (nRBCs) and microorganisms are also present in peripheral blood, which are of clinical
significance and important for disease diagnosis and fundamental research. As these cells usually occur
in low abundance (~1–10,000 cells/ mL of blood (~109 RBCs)), common macroscale separation techniques
such as physical filtration, fluorescence-activated cell sorting (FACS) and magnetically-actuated cell
sorting (MACS) have limited success in separating/isolating them mainly due to laborious sample
preparations which introduce artifacts or lead to loss of desired cells [68]. Moreover, mechanical
stress-induced changes in original cell phenotype using these techniques have also been reported,
suggesting the need to develop new tools for separating these rare cells [69,70]. Microfluidic based
separation methods are an attractive alternative due to the small length scale and numerous advantages
such as reduced sample and reagent volumes, minimize contamination issues and process
automation [4,34]. These inherent advantages also make microfluidics ideal in resource-poor settings
where problems such as shortage of power supply and refrigeration are limiting the use of expensive
and specialized equipment. For example, several microfluidic devices have been developed for malaria
detection, which is one of the severest parasitic blood-borne diseases and most prevalent in the poorer
countries. The separation principles are mostly based on intrinsic properties of malaria-infected RBCs
(iRBCs) such as dielectric differences [71], paramagnetic affinity [72], and change in stiffness [73,74]
to separate or enrich them from blood for diagnostic applications. However, as iRBCs constitute
~0.01–1% in peripheral blood of malaria patients, they are not considered as rare-cell per se and thus
excluded in our discussion. In this section, separation of various types of rare cells from blood using
microfluidics will be presented.
5.1. Nucleated RBCs
Prenatal diagnosis for genetic abnormalities in fetus is usually performed by collecting fetal
samples through amniocentesis and chorionic villus sampling (CVS), which are invasive procedures
and puts the patient at risk of miscarriage. While ultrasonography and serum biochemistry testing are
currently used to screen patients for non invasive prenatal diagnosis, it has been reported fetal
nucleated RBCs (nRBCs) present in maternal circulation can be used for direct analysis of fetal
chromosomes and their enumeration is also associated with fetal aneuploidies or pregnancy
complications [75,76]. In addition, nRBCs have short life span and are unlikely to persist between
pregnancies, making them a suitable candidate for direct fetal karyotype analysis. As nRBCs are
extremely rare (~1–30 nRBCs/mL of blood), enrichment or isolation of these cells usually consists of
density gradient centrifugation to obtain the mononuclear cells, followed by immuno-labeling and
FACS or MACS selection. These procedures are time consuming, labor-intensive and leads to
significant loss of nRBCs.
Several microfluidics devices have been developed to isolate these nRBCs using intrinsic markers
such as size and deformability, and hemoglobin content. The nRBCs are around 9–12 µm and similar
in size with leukocytes (10–15 µm), but larger than bulk of the RBCs (~6–8 µm in diameter).
Mohamed et al. have developed a simple microdevice which consists of an array of micropillars of
Micromachines 2011, 2
334
different gap sizes (2.5 to 15 µm) to separate the fetal nRBCs from blood based on size and
deformability [Figure 7(A)] [77]. Using mononuclear cells as their sample, they demonstrated that all
the leukocytes were trapped at the 2.5 µm width gaps while nRBCs deformed through the entire device
and were collected for subsequent DNA analysis. No immunolabeling was required in this process but
sample preparation including sample dilution and ficoll centrifugation were used to obtain the
mononuclear layer. The device throughput was also considerably low (~0.35 mL/h) taking ~6–8 h to
process each sample. To improve the throughput, Lee et al. have recently developed a microfluidic
device which continuously separates nRBCs from RBCs based on cross-flow filtration [78]. In this
system, sample and buffer solutions were pumped into the bottom and upper inlet respectively. Smaller
RBCs were filtered through the inclined filter of 4 µm gap and collected at the bottom outlet (port 4).
The bigger nRBCs, which could not deform through, would roll along the filter and be diverted
upstream into the upper outlet (port 3), thus achieving separation. Throughput was higher in this device
(0.3 mL/min) with good nRBCs recovery (74%), while 46.5% of RBCs were depleted in the upper
outlet (~1.7 fold enrichment). However, a major limitation was the absence of whole blood testing, in
which the bigger and stiffer leukocytes would probably clog the filters or be diverted upstream, resulting
in lower nRBCs purity. Moreover, in both aforementioned methods, only dilute blood samples could be
used as RBCs clog the filters easily which would affect the separation efficiency of nRBCs.
Recently, Huang et al. developed a 2-step microfluidic based isolation technique for nRBCs using
the popular ‘deterministic lateral displacement’ phenomenon [79]. In the first step, using the bump
array microchip, the larger leukocytes and nRBCs were separated from the RBCs and platelets. Post
sorting, the mononuclear cells (nRBCs and leukocytes) were treated with sodium nitrite to convert the
hemoglobin present in nRBCs to methemoglobin which is paramagnetic in nature allowing the
isolation of nRBCs using a magnetic column. This microfluidic-magnetic combination method was
able to eliminate 99.99% RBCs and leukocytes with 100% nRBCs identification in 58 predetermined
samples [79]. Throughput was also higher in this device (~2.5 mL/h) and whole blood was used
directly without sample dilution. Further work will be necessary to integrate both processes into a
single device to improve nRBCs isolation from blood for non invasive prenatal diagnosis.
5.2. Circulating Tumor Cells
Circulating tumor cells (CTCs) in blood is an important intermediate step in cancer metastasis,
which accounts for ~90% of all cancer related deaths [80]. Clinical studies have shown that
enumeration of CTCs can be correlated with cancer progression [81,82] and this “liquid biopsy” is
more practical as they can be performed more frequently than traditional invasive biopsies to predict
prognosis and guide therapy in cancer patients [83,84]. For CTCs enumeration to be useful in clinical
management of cancer, a highly efficient and robust CTC separation tool is required due to the low
abundance of these CTCs in blood. In the past decade, several microfluidics platforms have been
developed for CTCs separation from blood based on either physical filtration or affinity capture.
Due to the large size differences between CTCs and other blood cellular components (CTCs ~16–20 µm;
RBC ~8 µm; leukocytes ~10–15 µm) [85,86], size-based CTCs separation in microfluidic devices is
often achieved using microstructures of different geometries to physically trap the larger CTCs [87-90].
The critical dimension in these devices is the size of the pore/structure which is designed to allow
Micromachines 2011, 2
335
maximum CTCs trapping with little contamination. Using gap sizes between 5 to 10 µm, these
microfluidic devices are able to demonstrate isolation of cancer cells with high efficiency (>80%), but
the isolation purity varies greatly due to trapping of the bigger leukocytes [87-89]. The trapped CTCs
usually deforms at the pores under constant shear during sample processing, making their retrieval
difficult and thus requiring post-sorting analysis to be done on chip. To address this problem, Tan et al.
used an array of crescent-shaped isolation traps to isolate the CTCs in a more gentle manner which
allowed subsequent retrieval out of chip for downstream analysis [Figure 7(B)] [90]. The separation
principle is simple, does not require the use of biomarkers like EpCAM anti-bodies and allow the
retrieval of viable CTCs for further molecular characterization and analysis.
Apart from physical filtration methods, affinity based capture of CTCs in microfluidic devices
either by immobilizing specific capture molecules on magnetic beads [91], microstructures [92,93], or
array of channels [94] has been a popular CTC isolation approach. Using an array of anti EpCAM-coated
microposts, Nagrath et al. demonstrated that they were able to isolate CTCs from peripheral blood in
115 out of 116 (~99%) cancer patients in a single step [Figure 7(C)] [93]. However, a major drawback
for affinity capture is difficulty in finding ligands that have high specificity and affinity for CTCs due
to their heterogeneity in surface molecular signatures. The use of common epithelial markers such as
EpCAM is less desirable as EpCAM expression varies widely for different or even same tumor
type [95] and retrieval of isolated CTCs may be difficult due to cell binding in the devices. Moreover,
there is a risk of losing the most aggressive CTCs subpopulation due to epithelial to mesenchymal
transition (EMT) process, which leads to a down-regulation of epithelial surface markers and thus the
inability for CTCs to bind efficiently [96,97]. Recently, it has been demonstrated that aptamers
(single-stranded nucleic acid oligomers) could be used as probes to isolate CTCs with high affinity and
specificity [98,99]. As aptamers are synthesized by an in vitro process known as cell-SELEX
(Systematic Evolution of Ligands by EXponential enrichment), the surface molecular information of
CTCs are not required and different aptamer probes can be generated for any type or subpopulation of
CTCs. Whole blood can be used directly without dilution for affinity capture, but throughput is still
slow due to low shear rates to allow sufficient time for the CTCs to bind to capture surfaces as well as
to prevent bound CTCs from being detached under high shear.
5.3. Microorganisms
Separation of microorganisms from blood is primarily used as a therapeutic blood cleansing method
for treating sepsis, a lethal disease caused by a systemic microbial infection that spreads via the
bloodstream and compromise the body’s immunological system [100]. Antibiotics are often the main
course of treatment, but inflammation continues to spread for a few hours before the drugs are
effective. For newborn infants with sepsis (neonatal sepsis), their immune system are still functionally
immature and delay or inappropriate antibiotic treatment may cause fatality [101]. In this section we
will review several microfluidic technologies which have been developed to separate microorganisms
and pathogens from blood for potential sepsis treatment.
Micromachines 2011, 2
336
Figure 7. Rare cells separation from blood using microfluidics. (A) Schematic of the cross
flow filtration principle for nRBCs separation. RBCs are continuously filtered out through
the inclined filters with a 4 µm gap while the bigger nRBCs cannot deform through the
filters and are made to roll along the filter upstream and collected separately. Figure
reprinted from [78], copyright 2010, with kind permission from Elsevier; (B) Schematic
design of a microfluidic device for CTCs isolation and enumeration. Placing crescent-shaped
cell traps in the isolation regions traps the larger tumor cells while allowing smaller blood
components to sieve through. Figure reproduced from [90], copyright 2009, with
permission from Springer Science + Business Media, LLC; (C) Layout of the CTC chip
designed by Nagrath et al. consisting of functionalised micropost array for affinity capture
of CTCs. Inset showing a scanning electron micrograph of a lung cancer cell captured
between two microposts. Reprinted with permission from Macmillan Publishers Ltd:
Nature [93], copyright 2007; (D) Schematic illustrating the separation principle using a
micromagnetic-microfluidic based blood cleansing device to selectively remove pathogens
from contaminated blood. Immunomagnetic microbeads bind selectively to the pathogens
(see inset) to create magnetic opsonins which are removed continuously in the presence of
magnetic field gradients. Figure adapted with permission from [105], copyright 2009, The
Royal Society of Chemistry.
Typically microorganisms are very small ranging from 1 to 3 µm in diameter and thus can be easily
separated from the larger RBCs and leukocytes. Wu et al. separated bacteria from blood using inertial
lift force-induced migration based on size difference between bacteria and blood cells [102]. The
technique relies on the lateral displacement of the larger RBCs due to stronger wall-induced inertial lift
forces while the bacteria would remain in sample stream. The authors demonstrated removal of
~60% bacteria from a 10% (v/v) blood samples at a flow rate of 18 µL/min. Using similar inertial
focusing principle, Mach et al. developed a high throughput blood filtration device which consists of
Micromachines 2011, 2
337
40 high-aspect-ratio channels in parallel that can filter out bacteria from diluted blood sample
(0.5% hematocrit) at 8 mL/min continuously [103]. As inertial separation relies entirely on intrinsic
hydrodynamic forces, the design and device fabrication of these devices is simple and the large
channel dimensions also eliminate problems of channel clogging [4]. However, cell–cell interactions
can deteriorate the efficiency of this separation technique whereby only dilute samples can be used,
thus requiring additional sample preparation procedures.
Recently micromagnetic based microfluidic blood cleansing have been developed by Xia and
co-workers. The design consists of a microfluidic “H-filter” for continuous separation of E-coli from
blood using magnetic nanoparticles specifically bound to bacteria [Figure 7(D)] [104]. By fabricating a
magnetic (NiFe) microcomb structure adjacent to the channel, a steep magnetic field gradient can be
generated across the flow channel when magnetised externally allowing the opsonized bacteria to be
deflected from the original flow path and collected separately. The group demonstrated a throughput of
~104 cells/sec (at flow rates of 25 µL/h) with bacteria separation efficiency of ~78% using whole blood
sample spiked with E coli. The group later improved the design using an array of channels and
demonstrated fungal separation from whole blood with 1000× higher throughput (~20 mL/h) [105].
This method is highly versatile and can be used to separate multiple types of infectious microorganisms
simultaneously by using appropriately labeled magnetic beads. However, one would require prior
information on the pathogens to select the binding molecules for coating of magnetic beads.
6. Conclusions and Outlook
Here, we highlight the development of several microfluidic approaches that exploit the different
physical characteristics of cells, fluid mechanics and biorheology of blood as well as the difference in
binding affinity of specific targeted cells to separate and/or enrich them. These include red cells,
leukocytes, platelets, rare cells (such as CTCs) and pathogens (such as bacteria).
We envisage that we will continue to push the boundary of this research as more creative and
innovative methods continue to be developed and introduced in the near future. It is likely that they
will cater not only for the processing of both large and small volumes of blood, but are also fast and
efficient and with high sensitivity, purity and yield. However, for these blood-based separation systems
to be realized as true point-of-care (POC) devices, these techniques and devices must be simple, easy
to use and with little or no sample preparation required. This will set the challenges for researchers
working on POC devices and will by no means be easy, but possible.
References and Notes
1.
2.
3.
4.
5.
Whitesides, G.M. The origins and the future of microfluidics. Nature 2006, 442, 368-373.
Squires, T.M.; Quake, S.R. Microfluidics: Fluid physics at the nanoliter scale. Rev. Mod. Phys.
2005, 77, 977-1026.
Toner, M.; Irimia, D. Blood-on-a-chip. Annu. Rev. Biomed. Eng. 2005, 7, 77-103.
Bhagat, A.A.S.; Bow, H.; Hou, H.W.; Tan, S.; Han, J.; Lim, C.T. Microfluidics for cell
separation. Med. Biol. Eng. Comput. 2010, 48, 999-1014.
Pamme, N. Continuous flow separations in microfluidic devices. Lab Chip 2007, 7, 1644-1659.
Micromachines 2011, 2
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
338
Kersaudy-Kerhoas, M.; Dhariwal, R.; Desmulliez, M.P.Y. Recent advances in microparticle
continuous separation. IET Nanobiotechnol. 2008, 2, 1-13.
Eijkel, J.C.T.; van den Berg, A. Nanotechnology for membranes, filters and sieves. Lab Chip
2006, 6, 19-23.
Tsutsui, H.; Ho, C.M. Cell separation by non-inertial force fields in microfluidic systems.
Mech. Res. Commun. 2009, 36, 92-103.
Kulrattanarak, T.; van der Sman, R.G.M.; Schroen, C.; Boom, R.M. Classification and evaluation
of microfluidic devices for continuous suspension fractionation. Adv. Colloid Interface Sci. 2008,
142, 53-66.
Azim, W.; Azim, S.; Ahmed, K.; Shafi, H.; Rafi, T.; Luqman, M. Diagnostic significance of
serum protein electrophoresis. Biomedica 2004, 20, 40-44.
Kagan, B.M. The clinical significance of the serum protein’s. South. Med. J. 1943, 36, 234.
Haeberle, S.; Brenner, T.; Zengerle, R.; Ducrée, J. Centrifugal extraction of plasma from whole
blood on a rotating disk. Lab Chip 2006, 6, 776-781.
Lai, S.; Wang, S.; Luo, J.; Lee, L.J.; Yang, S.T.; Madou, M.J. Design of a compact disk-like
microfluidic platform for enzyme-linked immunosorbent assay. Anal. Chem. 2004, 76, 1832-1837.
Wong, A.P.; Gupta, M.; Shevkoplyas, S.S.; Whitesides, G.M. Egg beater as centrifuge: Isolating
human blood plasma from whole blood in resource-poor settings. Lab Chip 2008, 8, 2032-2037.
Brody, J.P.; Osborn, T.D.; Forster, F.K.; Yager, P. A planar microfabricated fluid filter.
Sens. Actuat. A 1996, 54, 704-708.
VanDelinder, V.; Groisman, A. Separation of plasma from whole human blood in a continuous
cross-flow in a molded microfluidic device. Anal. Chem. 2006, 78, 3765-3771.
Crowley, T.A.; Pizziconi, V. Isolation of plasma from whole blood using planar microfilters for
lab-on-a-chip applications. Lab Chip 2005, 5, 922-929.
Lin, L.; Guthrie, J.T. Preparation and characterisation of novel, blood-plasma-separation
membranes for use in biosensors. J. Membr. Sci. 2000, 173, 73-85.
Moorthy, J.; Beebe, D.J. In situ fabricated porous filters for microsystems. Lab Chip 2003, 3,
62-66.
Shim, J.S.; Browne, A.W.; Ahn, C.H. An on-chip whole blood/plasma separator with
bead-packed microchannel on COC polymer. Biomed. Microdevices 2010, 12, 949-957.
Yang, S.; Undar, A.; Zahn, J.D. A microfluidic device for continuous, real time blood plasma
separation. Lab Chip 2006, 6, 871-880.
Yang, S.; Ji, B.; Ündar, A.; Zahn, J.D. Microfluidic devices for continuous blood plasma
separation and analysis during pediatric cardiopulmonary bypass procedures. ASAIO J. 2006, 52,
698.
Fan, R.; Vermesh, O.; Srivastava, A.; Yen, B.K.H.; Qin, L.D.; Ahmad, H.; Kwong, G.A.;
Liu, C.C.; Gould, J.; Hood, L.; et al. Integrated barcode chips for rapid, multiplexed analysis of
proteins in microliter quantities of blood. Nat. Biotechnol. 2008, 26, 1373-1378.
Jaggi, R.D.; Sandoz, R.; Effenhauser, C.S. Microfluidic depletion of red blood cells from whole
blood in high-aspect-ratio microchannels. Microfluid. Nanofluid. 2007, 3, 47-53.
Blattert, C.; Jurischka, R.; Schoth, A.; Kerth, P.; Menz, W. Separation of blood cells and plasma
in microchannel bend structures. Proc. SPIE 2004, 5591, 143-151.
Micromachines 2011, 2
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
339
Blattert, C.; Jurischka, R.; Tahhan, I.; Schoth, A.; Kerth, P.; Menz, W. Separation of blood in
microchannel bends. Proc. SPIE 2004, 26, 2627-2630.
Lenshof, A.; Ahmad-Tajudin, A.; Jaras, K.; Sward-Nilsson, A.M.; Aberg, L.; Marko-Varga, G.;
Malm, J.; Lilja, H.; Laurell, T. Acoustic whole blood plasmapheresis chip for prostate specific
antigen microarray diagnostics. Anal. Chem. 2009, 81, 6030-6037.
Petersson, F.; Nilsson, A.; Holm, C.; Jönsson, H.; Laurell, T. Separation of lipids from blood
utilizing ultrasonic standing waves in microfluidic channels. Analyst 2004, 129, 938-943.
Petersson, F.; Nilsson, A.; Holm, C.; Jönsson, H.; Laurell, T. Continuous separation of lipid
particles from erythrocytes by means of laminar flow and acoustic standing wave forces.
Lab Chip 2005, 5, 20-22.
Yeo, L.Y.; Friend, J.R.; Arifin, D.R. Electric tempest in a teacup: The tea leaf analogy to
microfluidic blood plasma separation. Appl. Phys. Lett. 2006, 89, 103516.
Arifin, D.R.; Yeo, L.Y.; Friend, J.R. Microfluidic blood plasma separation via bulk
electrohydrodynamic flows. Biomicrofluidics 2007, 1, 014103.
Xu, C.X.; Wang, Y.; Cao, M.; Lu, Z.H. Dielectrophoresis of human red cells in microchips.
Electrophoresis 1999, 20, 1829-1831.
Nakashima, Y.; Hata, S.; Yasuda, T. Blood plasma separation and extraction from a minute
amount of blood using dielectrophoretic and capillary forces. Sens. Actuat. B 2010, 145, 561-569.
Pratt, E.D.; Huang, C.; Hawkins, B.G.; Gleghorn, J.P.; Kirby, B.J. Rare cell capture in
microfluidic devices. Chem. Eng. Sci. 2010, 66, 1508-1522.
Buckner, D.; Graw, R.G., Jr.; Eisel, R.J.; Henderson, E.S.; Perry, S. Leukapheresis by continuous
flow centrifugation (CFC) in patients with chronic myelocytic leukemia (CML). Blood 1969, 33,
353.
Malachowski, M.E.; Comenzo, R.L.; Hillyer, C.D.; Tiegerman, K.O.; Berkman, E.M. Large
volume leukapheresis for peripheral blood stem cell collection in patients with hematologic
malignancies. Transfusion 1992, 32, 732-735.
Chu, R.W. Leukocytes in blood transfusion: Adverse effects and their prevention. Hong Kong
Med. J. 1999, 5, 280-284.
Schmid-Schonbein, G.W.; Shih, Y.Y.; Chien, S. Morphometry of human leukocytes. Blood
1980, 56, 866-875.
Downey, G.P.; Doherty, D.E.; Schwab 3rd, B.; Elson, E.L.; Henson, P.M.; Worthen, G.S.
Retention of leukocytes in capillaries: Role of cell size and deformability. J. Appl. Physiol. 1990,
69, 1767-1778.
Sethu, P.; Sin, A.; Toner, M. Microfluidic diffusive filter for apheresis (leukapheresis). Lab Chip
2006, 6, 83-89.
Bruil, A.; Van Aken, W.G.; Beugeling, T.; Feijen, J.; Steneker, I.; Huisman, J.G.; Prins, H.K.
Asymmetric membrane filters for the removal of leukocytes from blood. J. Biomed. Mater. Res.
1991, 25, 1459-1480.
Kraus, M.; Yonath, J. Leukocyte Removal Method Using A Nitrocellulose Membrane Filter
Unit. US Patent 5707526, 1998.
Panaro, N.J.; Lou, X.J.; Fortina, P.; Kricka, L.J.; Wilding, P. Micropillar array chip for integrated
white blood cell isolation and PCR. Biomol. Eng. 2005, 21, 157-162.
Micromachines 2011, 2
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
340
Ji, H.M.; Samper, V.; Chen, Y.; Heng, C.K.; Lim, T.M.; Yobas, L. Silicon-based microfilters for
whole blood cell separation. Biomed. Microdevices 2008, 10, 251-257.
VanDelinder, V.; Groisman, A. Perfusion in microfluidic cross-flow: Separation of white blood
cells from whole blood and exchange of medium in a continuous flow. Anal. Chem. 2007, 79,
2023-2030.
Huang, L.R.; Cox, E.C.; Austin, R.H.; Sturm, J.C. Continuous particle separation through
deterministic lateral displacement. Science 2004, 304, 987-990.
Davis, J.A.; Inglis, D.W.; Morton, K.J.; Lawrence, D.A.; Huang, L.R.; Chou, S.Y.; Sturm, J.C.;
Austin, R.H. Deterministic hydrodynamics: Taking blood apart. Proc. Natl. Acad. Sci. USA
2006, 103, 14779-14784.
Yamada, M.; Seki, M. Hydrodynamic filtration for on-chip particle concentration and
classification utilizing microfluidics. Lab Chip 2005, 5, 1233-1239.
Zheng, S.; Liu, J.Q.; Tai, Y.C. Streamline-based microfluidic devices for erythrocytes and
leukocytes separation. J. Microelectromech. Syst. 2008, 17, 1029-1038.
Choi, S.; Song, S.; Choi, C.; Park, J.K. Continuous blood cell separation by hydrophoretic
filtration. Lab Chip 2007, 7, 1532-1538.
Shevkoplyas, S.S.; Yoshida, T.; Munn, L.L.; Bitensky, M.W. Biomimetic autoseparation of
leukocytes from whole blood in a microfluidic device. Anal. Chem. 2005, 77, 933-937.
Jain, A.; Munn, L.L. Determinants of leukocyte margination in rectangular microchannels.
PLoS One 2009, 4, e7104.
Chang, W.C.; Lee, L.P.; Liepmann, D. Biomimetic technique for adhesion-based collection and
separation of cells in a microfluidic channel. Lab Chip 2005, 5, 64-73.
Goldsmith, H.L.; Spain, S. Margination of leukocytes in blood flow through small tubes.
Microvasc. Res. 1984, 27, 204-222.
SooHoo, J.R.; Walker, G.M. Microfluidic aqueous two phase system for leukocyte concentration
from whole blood. Biomed. Microdevices 2009, 11, 323-329.
Sethu, P.; Moldawer, L.L.; Mindrinos, M.N.; Scumpia, P.O.; Tannahill, C.L.; Wilhelmy, J.;
Efron, P.A.; Brownstein, B.H.; Tompkins, R.G.; Toner, M. Microfluidic isolation of leukocytes
from whole blood for phenotype and gene expression analysis. Anal. Chem. 2006, 78, 5453-5461.
Han, K.H.; Frazier, A.B. Lateral-driven continuous dielectrophoretic microseparators for blood
cells suspended in a highly conductive medium. Lab Chip 2008, 8, 1079-1086.
Inglis, D.W.; Riehn, R.; Austin, R.H.; Sturm, J.C. Continuous microfluidic immunomagnetic cell
separation. Appl. Phys. Lett. 2004, 85, 5093-5095.
Inglis, D.W.; Riehn, R.; Sturm, J.C.; Austin, R.H. Microfluidic high gradient magnetic cell
separation. J. Appl. Phys. 2006, 99, 08K101.
Qu, B.-Y.; Wu, Z.-Y.; Fang, F.; Bai, Z.-M.; Yang, D.-Z.; Xu, S.-K. A glass microfluidic chip for
continuous blood cell sorting by a magnetic gradient without labeling. Anal. BioAnal. Chem.
2008, 392, 1317-1324.
Han, K.H.; Frazier, A.B. Paramagnetic capture mode magnetophoretic microseparator for high
efficiency blood cell separations. Lab Chip 2006, 6, 265-273.
Micromachines 2011, 2
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
341
Basabe-Desmonts, L.; Ramstrom, S.; Meade, G.; O'Neill, S.; Riaz, A.; Lee, L.P.; Ricco, A.J.;
Kenny, D. Single-step separation of platelets from whole blood coupled with digital
quantification by interfacial platelet cytometry (iPC). Langmuir 2010, 26, 14700-14706.
Triulzi, D.J.; Griffith, B.P. Blood usage in lung transplantation. Transfusion 1998, 38, 12-15.
Perrotta, P.L.; Synder, E.L. Platelet storage and transfusion. In Platelets; Michelson, A.D., Ed.;
Science: Amsterdam, the Netherlands, 2007; pp. 1265-1295.
Brass, L.F.; Stalker, T.J.; Zhu, L.; Woulfe, D.S. Signal transduction during platelet plug
formation. In Platelets; Michelson, A.D., Ed.; Burlington: Academic Press, 2007; pp. 319-346.
Petersson, F.; Aberg, L.; Sward-Nilsson, A.M.; Laurell, T. Free flow acoustophoresis:
Microfluidic-based mode of particle and cell separation. Anal. Chem. 2007, 79, 5117-5123.
Pommer, M.S.; Zhang, Y.T.; Keerthi, N.; Chen, D.; Thomson, J.A.; Meinhart, C.D.; Soh, H.T.
Dielectrophoretic separation of platelets from diluted whole blood in microfluidic channels.
Electrophoresis 2008, 29, 1213-1218.
Dharmasiri, U.; Witek, M.A.; Adams, A.A.; Soper, S.A. Microsystems for the capture of
low-abundance cells. Annu. Rev. Anal. Chem. 2010, 3, 409-431.
Lundahl, J.; Hallden, G.; Hallgren, M.; Skold, C.M.; Hed, J. Altered expression of CD11b/CD18
and CD62L on human monocytes after cell preparation procedures. J. Immunol. Methods 1995,
180, 93-100.
Fukuda, S.; Schmid-Schonbein, G.W. Centrifugation attenuates the fluid shear response of
circulating leukocytes. J. Leukocyte Biol. 2002, 72, 133-139.
Gascoyne, P.; Mahidol, C.; Ruchirawat, M.; Satayavivad, J.; Watcharasit, P.; Becker, F.F.
Microsample preparation by dielectrophoresis: Isolation of malaria. Lab Chip 2002, 2, 70-75.
Zimmerman, P.A.; Thomson, J.M.; Fujioka, H.; Collins, W.E.; Zborowski, M. Diagnosis of
malaria by magnetic deposition microscopy. Am. J. Trop. Med. Hyg. 2006, 74, 568-572.
Hou, H.W.; Bhagat, A.A.S.; Chong, A.G.L.; Mao, P.; Tan, K.S.W.; Han, J.Y.; Lim, C.T.
Deformability based cell margination—A simple microfluidic design for malaria-infected
erythrocyte separation. Lab Chip 2010, 10, 2605-2613.
Bow, H.; Pivkin, I.V.; Diez-Silva, M.; Goldfless, S.J.; Dao, M.; Niles, J.C.; Suresh, S.; Han, J.
A microfabricated deformability-based flow cytometer with application to malaria. Lab Chip
2011, 11, 1065-1073.
Mavrou, A.; Kouvidi, E.; Antsaklis, A.; Souka, A.; Tzeli, S.K.; Kolialexi, A. Identification of
nucleated red blood cells in maternal circulation: A second step in screening for fetal
aneuploidies and pregnancy complications. Prenatal Diagn. 2007, 27, 150-153.
Bianchi, D.W.; Simpson, J.L.; Jackson, L.G.; Elias, S.; Holzgreve, W.; Evans, M.I.; Dukes, K.A.;
Sullivan, L.M.; Klinger, K.W.; Bischoff, F.Z.; et al. Fetal gender and aneuploidy detection using
fetal cells in maternal blood: Analysis of NIFTY I data. Prenatal Diagn. 2002, 22, 609-615.
Mohamed, H.; Turner, J.N.; Caggana, M. Biochip for separating fetal cells from maternal
circulation. J. Chromatogr. A 2007, 1162, 187-192.
Lee, D.; Sukumar, P.; Mahyuddin, A.; Choolani, M.; Xu, G. Separation of model mixtures of
epsilon-globin positive fetal nucleated red blood cells and anucleate erythrocytes using a
microfluidic device. J. Chromatogr. A 2010, 1217, 1862-1866.
Micromachines 2011, 2
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
342
Huang, R.; Barber, T.A.; Schmidt, M.A.; Tompkins, R.G.; Toner, M.; Bianchi, D.W.; Kapur, R.;
Flejter, W.L. A microfluidics approach for the isolation of nucleated red blood cells (NRBCs)
from the peripheral blood of pregnant women. Prenatal Diagn. 2008, 28, 892-899.
Wittekind, C.; Neid, M. Cancer invasion and metastasis. Oncology 2005, 69, 14-16.
Budd, G.T.; Cristofanilli, M.; Ellis, M.J.; Stopeck, A.; Borden, E.; Miller, M.C.; Matera, J.;
Repollet, M.; Doyle, G.V.; Terstappen, L.; et al. Circulating tumor cells versus imaging-predicting
overall survival in metastatic breast cancer. Clin. Cancer Res. 2006, 12, 6403-6409.
Cristofanilli, M.; Budd, G.T.; Ellis, M.J.; Stopeck, A.; Matera, J.; Miller, M.C.; Reuben, J.M.;
Doyle, G.V.; Allard, W.J.; Terstappen, L. Circulating tumor cells, disease progression, and
survival in metastatic breast cancer. N. Engl. J. Med. 2004, 351, 781-791.
Kaiser, J. Cancer’s circulation problem. Science 2010, 327, 1072-1074.
Hayes, D.F.; Cristofanilli, M.; Budd, G.T.; Ellis, M.J.; Stopeck, A.; Miller, M.C.; Matera, J.;
Allard, W.J.; Doyle, G.V.; Terstappen, L. Circulating tumor cells at each follow-up time point
during therapy of metastatic breast cancer patients predict progression-free and overall survival.
Clin. Cancer Res. 2006, 12, 4218-4224.
Vona, G.; Sabile, A.; Louha, M.; Sitruk, V.; Romana, S.; Schutze, K.; Capron, F.; Franco, D.;
Pazzagli, M.; Vekemans, M.; et al. Isolation by size of epithelial tumor cells—A new method for
the immunomorphological and molecular characterization of circulating tumor cells. Am. J.
Pathol. 2000, 156, 57-63.
Zabaglo, L.; Ormerod, M.G.; Parton, M.; Ring, A.; Smith, I.E.; Dowsett, M. Cell filtration-laser
scanning cytometry for the characterisation of circulating breast cancer cells. Cytometry, Part A
2003, 55A, 102-108.
Mohamed, H.; Murray, M.; Turner, J.N.; Caggana, M. Isolation of tumor cells using size and
deformation. J. Chromatogr. A 2009, 1216, 8289-8295.
Zheng, S.; Lin, H.; Liu, J.-Q.; Balic, M.; Datar, R.; Cote, R.J.; Tai, Y.-C. Membrane microfilter
device for selective capture, electrolysis and genomic analysis of human circulating tumor cells.
J. Chromatogr. A 2007, 1162, 154-161.
Hosokawa, M.; Hayata, T.; Fukuda, Y.; Arakaki, A.; Yoshino, T.; Tanaka, T.; Matsunaga, T.
Size-selective microcavity array for rapid and efficient detection of circulating tumor cells.
Anal. Chem. 2010, 82, 6629-6635.
Tan, S.J.; Yobas, L.; Lee, G.; Ong, C.; Lim, C.T. Microdevice for the isolation and enumeration
of cancer cells from blood. Biomed. Microdevices 2009, 11, 883-892.
Saias, L.; Saliba, A.E.; Viovy, J.L.; Pierga, J.Y.; Vielh, P.; Farace, F. Microfluidic magnetic cell
sorting system for cancer diagnosis. Houille Blanche 2009, 5, 105-111.
Gleghorn, J.P.; Pratt, E.D.; Denning, D.; Liu, H.; Bander, N.H.; Tagawa, S.T.; Nanus, D.M.;
Giannakakou, P.A.; Kirby, B.J. Capture of circulating tumor cells from whole blood of prostate
cancer patients using geometrically enhanced differential immunocapture (GEDI) and a
prostate-specific antibody. Lab Chip 2010, 10, 27-29.
Nagrath, S.; Sequist, L.V.; Maheswaran, S.; Bell, D.W.; Irimia, D.; Ulkus, L.; Smith, M.R.;
Kwak, E.L.; Digumarthy, S.; Muzikansky, A.; et al. Isolation of rare circulating tumour cells in
cancer patients by microchip technology. Nature 2007, 450, 1235-1239.
Micromachines 2011, 2
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
343
Adams, A.A.; Okagbare, P.I.; Feng, J.; Hupert, M.L.; Patterson, D.; Gottert, J.; McCarley, R.L.;
Nikitopoulos, D.; Murphy, M.C.; Soper, S.A. Highly efficient circulating tumor cell isolation
from whole blood and label-free enumeration using polymer-based microfluidics with an
integrated conductivity sensor. J. Am. Chem. Soc. 2008, 130, 8633-8641.
Spizzo, G.; Gastl, G.; Obrist, P.; Went, P.; Dirnhofer, S.; Bischoff, S.; Mirlacher, M.; Sauter, G.;
Simon, R.; Stopatschinskaya, S.; et al. High Ep-CAM expression is associated with poor
prognosis in node-positive breast cancer. Breast Cancer Res. Treat. 2004, 86, 207-213.
Lee, J.M.; Dedhar, S.; Kalluri, R.; Thompson, E.W. The epithelial-mesenchymal transition: New
insights in signaling, development, and disease. J. Cell Biol. 2006, 172, 973-981.
Thompson, E.W.; Newgreen, D.F.; Tarin, D. Carcinoma invasion and metastasis: A role for
epithelial-mesenchymal transition? Cancer Res. 2005, 65, 5991-5995.
Dharmasiri, U.; Balamurugan, S.; Adams, A.A.; Okagbare, P.I.; Obubuafo, A.; Soper, S.A.
Highly efficient capture and enumeration of low abundance prostate cancer cells using
prostate-specific membrane antigen aptamers immobilized to a polymeric microfluidic device.
Electrophoresis 2009, 30, 3289-3300.
Xu, Y.; Phillips, J.A.; Yan, J.L.; Li, Q.G.; Fan, Z.H.; Tan, W.H. Aptamer-based microfluidic
device for enrichment, sorting, and detection of multiple cancer cells. Anal. Chem. 2009, 81,
7436-7442.
Cohen, J. The immunopathogenesis of sepsis. Nature 2002, 420, 885-891.
Edmond, K.; Zaidi, A. New approaches to preventing, diagnosing, and treating neonatal sepsis.
PLoS Med. 2010, 7, e1000213.
Wu, Z.G.; Willing, B.; Bjerketorp, J.; Jansson, J.K.; Hjort, K. Soft inertial microfluidics for high
throughput separation of bacteria from human blood cells. Lab Chip 2009, 9, 1193-1199.
Mach, A.J.; Di Carlo, D. Continuous scalable blood filtration device using inertial microfluidics.
Biotechnol. Bioeng. 2010, 107, 302-311.
Xia, N.; Hunt, T.; Mayers, B.; Alsberg, E.; Whitesides, G.; Westervelt, R.; Ingber, D. Combined
microfluidic-micromagnetic separation of living cells in continuous flow. Biomedical
Microdevices 2006, 8, 299-308.
Yung, C.W.; Fiering, J.; Mueller, A.J.; Ingber, D.E. Micromagnetic-microfluidic blood cleansing
device. Lab Chip 2009, 9, 1171-1177.
© 2011 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article
distributed under the terms and conditions of the Creative Commons Attribution license
(http://creativecommons.org/licenses/by/3.0/).